Absolute cosine-based SVM-RFE feature selection method for prostate histopathological grading

作者：

Highlights：

• SVM-RFE(AC) feature selection method minimises the redundancy prevalent in SVM-RFE.

• SVM-RFE(AC) avoids overfitting in feature selection for histopathology images.

• SVM-RFE(AC) outperforms other methods in the challenging task of prostate grading.

摘要

•SVM-RFE(AC) feature selection method minimises the redundancy prevalent in SVM-RFE.•SVM-RFE(AC) avoids overfitting in feature selection for histopathology images.•SVM-RFE(AC) outperforms other methods in the challenging task of prostate grading.

论文关键词：Prostate histopathological image,Tissue components,Ensemble classification,Feature selection,SVM-RFE,Absolute cosine,Redundancy

论文评审过程：Received 29 April 2017, Revised 2 April 2018, Accepted 7 April 2018, Available online 19 April 2018, Version of Record 28 May 2018.

论文官网地址：https://doi.org/10.1016/j.artmed.2018.04.002